Conclusions - DISCUSSION AND CONCLUSIONS - Development of remote sensing techniques for the imp

CHAPTER 5 DISCUSSION AND CONCLUSIONS

5.6 Conclusions

This study, which investigated species reflectance differences, segmentation, and image classification techniques for discrimination of field grown crop/weed species combinations, shows potential for the use of remotely sensed data for high resolution herbicide prescription maps. Classification of these image data provides a means of mapping weed location and density within a field, an integral step in successful application of SSHM techniques.

Segmentation of foreground and background pixels from imagery was presented as an initial step in eliminating background pixels and in turn classifying only pixels of interest. Through visual analysis, the MCARI showed excellent results in terms of consistency in identification of vegetated pixels.

Classification using MLC and ANNs proved useful for discrimination of single weed/crop mixtures over both image acquisition dates. Generally the crop/RRP

classification of the PEA/RRP combination. The earlier plant stage (July 19) showed consistently better classification results than the latter July 26 acquisition, suggesting that optimal species discrimination can be obtained at early plant growth stages. Since observed reflectance characteristics varied as a function of plant stage, both classification techniques were evaluated for their ability to account for temporal variability. ANN models, with their efficiency in handling complex feature space, provided better results than MLC in the multitemporal series of classifications. Therefore creation of a single classification model trained to encompass the variability of spectral reflectance throughout the optimal herbicide application timeframe is possible.

The identification of important spectral bands for species discrimination was achieved through PCA and SDA with seven wavebands selected from the original 61 waveband dataset. This waveband reduction slightly lowered ANN overall accuracies (less than 3%), with most crop/weed combinations and suggested that a seven band multispectral sensor may be adequate in discrimination between two plant species. 5.7 Future Research

Very high resolution image data can be prone to spectral variability caused by leaf angle and orientation effects as well as atmospheric illumination changes. In this study, image data were collected only on clear sunny days through hours of peak incoming solar radiation to reduce the radiation variability. Typical application of herbicides to

agricultural systems would be throughout the entire day and would require constant calibration for accurate reflectance measurements throughout this period. Further research must focus on effects of differential illumination conditions as well as

growing within the same field was also an issue that was not examined and further study is required to address this scenario which represents real-world application of SSHM techniques.

The movement towards successful application of SSHM techniques will only be achieved through investigation of new emergent technologies. Though these

technologies are by no means cost effective for direct SSHM implementation, new procedures and evaluation of these advancements is essential to SSHM development. Investigation into species discrimination and application of remotely sensed data for mapping weed location and density today, leads to end-use application tomorrow. As imaging systems and processing technology become more widely available and cost effective with proven research behind them, acceptance and implementation will surely follow.

REFERENCES CITED

Aitkenhead, M.J., Dalgetty, I.A., Mullins, C.E., McDonald, A.J.S., and Strachan, N.J.S. (2003) Weed and crop discrimination using image analysis and artifical intelligence methods. Computers and Electronics in Agriculture, 39: 157-171.

Atkinson, P.M., and Tatnall, A.R.L. (1997) Neural networks in remote sensing.

International Journal of Remote Sensing, 18(4): 699-709.

Bechdol, M.A., Gualtieri, J.A., Hunt, J.T., Chettri, S., and Garegnani, J. (2000) Hyperspectral imaging: a potential tool for improving weed and herbicide

management. In Proc. 5th

Int. Conference on Precision Agriculture, Bloomington,

Minnesota, 16-19 July 2000.

Benediktsson, J.A., Swain, P.H. and Ersoy, O.K. (1990) Neural network approaches versus statistical methods in classification of multisource remote sensing data. IEEE

Transactions on Geoscience and Remote Sensing, 28:540-552.

Beucher, S. (1992) The watershed transformation applied to image segmentation. In 10th

Pfefferkorn Conference on Signal and Image Processing in Microscopy and Microanalysis, Cambridge, UK, 16-19 September, 1991.

Beucher, S., and Lantuejoul, C. (1979) Use of watersheds in contour detection. In

Proceedings International Workshop on Image Processing, Real-time Edge and Motion Detection/Estimation, CCETT/IRISA, France, pp. 180-186.

Bleau, A., and Leon, L.J. (2000) Watershed-based segmentation and region merging.

Computer Vision and Image Understanding, 77: 317-370.

Borregaard, T., Neilsen, H., Norgaard, L., and Have, H. (2000) Crop-weed discrimination by line imaging spectroscopy. Journal of Agricultural Engineering Research, 75: 389-400.

Brown, R.B., and Noble, S.D. (2005) Site-specific weed management: sensing requirements-what do we need to see? Weed Science, 53: 252-258.

Brown, R.B., and Steckler, J.-P.G.A. (1993) Weed patch identification in no till corn using digital imagery. Canadian Journal of Remote Sensing, 19(1): 88-91. Brown, R.B., and Steckler, J.-P.G.A. (1995) Prescription maps for spatially variable

herbicide application in no-till corn. Transactions of the ASAE, 38(6): 1659-1666. Brown, R.B., Steckler, J.-P.G.A., and Anderson, G.W. (1994) Remote sensing for

Brown, R.J., Staenz, K., McNairn, H., Hopp, B., and Van Acker, R. (1997) Application of high resolution optical imagery to precision agriculture. International Symposium,

Geomatics in the Era of RADARSAT (GER'97), Ottawa, Canada, 25-30 May, 1997,

pp. 9.

Burks, T.F., Shearer, S.A., Gates, R.S., and Donohue, K.D. (2000b) Backpropagation neural network design and evaluation for classifying weed species using color image texture. Transactions of the ASAE, 43(4): 1029-1037.

Burks, T.F., Shearer, S.A., Heath, J.R., and Donohue, K.D. (2005) Evaluation of neural- network classifiers for weed species discrimination. Biosystems Engineering, 91(3): 293-304.

Burks, T.F., Shearer, S.A., and Payne, F.A. (2000a) Classification of weed species using color texture features and discriminant analysis. Transactions of the ASAE, 43(2): 441-448.

Carter, G.A. (1991) Primary and secondary effects of water content on the spectral reflectance of leaves. American Journal of Botany, 78(7): 916-924.

Chapron, M., Requena-Esteso, M., Boissard, P., and Assemat, L. (1999) A method for recognizing vegetal species from multispectral images. In: Precision Agriculture

1999 2nd European Conference on Precision Agriculture (JV Stafford, Ed.),

Sheffield Academic Press, Sheffield, UK. pp. 239-248.

Cheng, H.D., Jiang, X.H., Sun, Y., and Wang, J. (2001) Color image segmentation: advances and prospects. Pattern Recognition, 34: 2259-2281.

Cho, S.I., Lee, D.S., and Jeong, J.Y. (2002) Weed-plant discrimination by machine vision and artificial neural network. Biosystems Engineering, 83(3): 275-280.

Crop Life (2007) Annual Report 2005/2006. Website:

http://www.croplife.ca/english/pdf/annnualreport/annualreport05-6-eng.pdf, Accessed April 4, 2007.

Curran, P.J. (1989) Remote sensing of foliar chemistry. Remote Sensing of Environment, 30: 271-278.

Daughtry, C.S.T., Walthall, C.L., Kim, M.S., Brown de Colstoun, E., and McMurtrey, J.E. (2000) Estimating corn leaf chlorophyll concentration from leaf canopy reflectance. Remote Sensing of Environment, 74: 229-239.

Edmund Optics (2007) Online Catalogue 2007. Website:

http://edmundoptics.com/onlineCatalog/displayproduct.cfm?productID=1615, Accessed May 29, 2007.

El-Faki, M.S., Zhang, N., and Peterson, D.E. (2000a) Factors affecting color-based weed detection. Transactions of the ASAE, 43(4): 1001-1009.

El-Faki, M.S., Zhang, N., and Peterson, D.E. (2000b) Weed detection using color machine vision. Transactions of the ASAE, 43(6): 1969-1978.

Foley, J.D. and Van Dam A. (1982) Fundamentals of interactive computer graphics. Addison-Wesley Longman Publishing Co. Inc., Boston, Massachusetts.

Gausman, H.W., Allen, W.A., Weigland, C.L., Escobar, D.E., and Richardson, A.J. (1973) The leaf mesophylls of twenty crops, their light spectra, and optical and geometrical parameters. USDA Technical Bulletin No. 1465, U.S. Government Printing office, Washington, D.C.

Grant, L. (1987) Diffuse and specular characteristics of leaf reflectance. Remote Sensing

of Environment, 22: 309-322.

Haboudane, D., Miller, J.R., Tremblay, N., Zarco-Tejada, P.J., and Dextraze, L. (2002) Integrated narrow-band vegetation indices for prediction of crop chlorophyll content for application to precision agriculture. Remote Sensing of Environment, 81: 416-426. Haralick, R.M., Sternberg, S.R., and Zhuang, X. (1987) Image analysis using

mathematical morphology. IEEE Transactions on Pattern Analysis and Machine

Intelligence, PAMI-9(4): 532-550.

Jensen, J.R. (1996) Introductory Digital Image Processing; A Remote Sensing

Pertspective. (K.C. Clarke Ed.), Prentice-Hall, Inc, Upper Saddle River, New Jersey.

Kanellopoulos, I., and Wilkinson, G.G. (1997) Strategies and best practice for neural network image classification. International Journal of Remote Sensing, 18(4): 711- 725.

Kavdir, I. (2004) Discrimination of sunflower, weed and soil by artificial neural networks. Computers and Electronics in Agriculture, 44: 153-160.

Key, J., Maslanic, A. and Schweiger, A.J., (1989) Classification of merged AVHRR and SMMR arctic data with neural networks. Photogrammetric Engineering and Remote

Sensing, 55: 1331-1338.

Lamb, D.W. (2000) The use of qualitative airborne multispectral imaging for managing agricultural crops – a case study in south-eastern Australia. Australian Journal of

Experimental Agriclture, 40: 725-738.

Lamb, D.W, and Brown, R.B. (2001) Remote-sensing and mapping of weeds in crops.

Lamb, D.W., and Weedon, M.M. (1998) Evaluating the accuracy of mapping weeds in fallow fields using airborne digital imaging: Panicum effusum in oilseed rape stubble.

Weed Research, 38: 443-451.

Lamb, D.W., Weedon, M.M., and Rew, L.J. (1999) Evaluating the accuracy of mapping weeds in seedling crops using airborne digital imaging: Avena spp. in seedling triticale. Weed Research, 39: 481-492.

Lawrence, J. (1994) Neural networks: design, theory and applications. (S. Luedeking and J. Fredrickson, Eds.), California Scientific Software, Nevada City, California. Lewis, M. (2002) Spectral characterization of Australian arid zone plants. Canadian

Journal of Remote Sensing, 28(2): 219-230.

Li, W., Benie, G.B., He, D., Wang, S., Ziou, D., and Gwyn, Q.H.J. (1999) Watershed- based hierarchical SAR image segmentation. International Journal of Remote

Sensing, 20: 3377-3390.

Lillesand, T.M., and Kiefer, R.W. (1987) Remote sensing and image interpretation. John Wiley and Sons, New York, New York.

Lindquist, J.L., Dieleman, J.A., Mortensen, D.A., Johnson, G.A., and Wyse-Pester D.Y. (1998) Economic importance of managing spatially heterogeneous weed populations.

Weed Technology, 12: 7-13.

Lippert, R.M., and Wolak, F.J. (1999) Weed mapping and assessment of broadcast vs.

spot treatment of sicklepod weeds in soybeans. 2nd

European Conference on Precision Agriculture, Odense, Denmark, 11-15 July, 1999.

Martin-Chefson, L., Chapron, M., and Philipp, S. (1999) A two dimensional method for recognising weeds from multi-band image processing. In: Precision Agriculture 1999

2nd European Conference on Precision Agriculture (J.V. Stafford, Ed.), Sheffield

Academic Press, Sheffield, UK. pp. 473-483.

Medlin, C.R., Shaw, D.R., Gerard, P.D., and LaMastus, F.E. (2000) Using remote sensing to detect weed infestations in Glycine max. Weed Science, 48: 393-398. Moshou, D., Vrindts, E., De Detelaere, B., De Baerdemaeker, J., and Ramon, H. (2001)

A neural network based plant classifier. Computers and Electronics in Agriculture, 31: 5-16.

Mortensen, D.A., Johnson, G.A., Wyse, D.Y., and Martin, A.R. (1995) Managing spatially variable weed populations. In Site-Specific Management for Agricultural

Systems, (P. Robert, R.H. Rust, and W.E. Larson ,Eds), ASA, CSSA, SSSA,

Mulla, D.J., and Schepers, J.S. (1997) Key processes and properties for site-specific soil and crop management. In Site-Specific Management for Agricultural Systems, (P. Robert, R.H. Rust, and W.E. Larson ,Eds), ASA, CSSA, SSSA, Madison, WI. pp. 1- 18.

Noble, S.D., Brown, R.B., and Crowe, T.G. (2002) The use of spectral properties for weed detection and identification – A review. Presented at AIC 2002 Meeting,

CSAR/SCGR Program, Saskatoon, Saskatchewan. Paper No. 02-208.

Noble, S.D., and Crowe, T.G. (2005) Analysis of crop and weed leaf diffuse reflectance spectra. Transactions of the ASAE, 48(6): 2379-2387.

Radhakrishnan, R., Teasdale, J.R., Liang, S., and Shuey, C.J. (2002) Remote sensing of weed canopies. In Laboratory Spectroscopy to Remotely Sensor Spectra of Terrestrial

Ecosystems (R.S Muttiah, Ed.), Kluwer Academic Publishers, Netherlands. pp. 175-

202.

Rambabu, C., Chakrabarti, I., and Mahanta, A. (2004) Flooding-based watershed

algorithm and its prototype hardware architecture. IEE Proceedings of Vision, Image

& Signal Processing, 151(3): 224-234.

Rosenblatt, F. (1958) The perceptron: a probabilistic model for information storage and organization in the brain. Psychological Review, 65:386-408.

Rosin, P.L. (2003) Measuring shape: ellipticity, rectangularity, and triangularity.

Machine Vision and Applications, 14: 172-184.

Ruberto, C.D., Dempster, A., Kahn, S., and Jarra, B. (2002) Analysis of infected blood cell images using morphological operators. Image and Vision Computing, 20: 133- 146.

Soille, P. (2000) Morphological image analysis applied to crop field mapping. Image and

Vision Computing, 18: 1025-1032.

Smith, A.M., and Blackshaw, R.E. (2003) Weed-crop discrimination using remote sensing: a detached leaf experiment. Weed Technology, 17: 811-820.

Smith, M., (1993) Neural networks for statistical modeling. Van Nostrand Reinhold, New York, New York.

Swinton, S.M. (2005) Economics of site-specific weed management. Weed Science, 53: 259-263.

Tang, L., Steward, B.L., and Tian, L.F. (1999) Experimental study of a machine vision- based real-time weed sensing patch-sprayer. Society of Automotive Engineers. Paper No. 99F-38(UIUC-ENG-99-7009).

Tang.L., Tian, L.F., and Steward, B.L. (2000) Color image segmentation with genetic algorithm for in-field weed sensing. Transactions of the ASAE. 43(4): 1019-1027. Thenkabail, P.S., Enclona, E.A., Ashton, M.S., and Van Der Meer, B. (2004) Accuracy

assessments of hyperspectral waveband performance for vegetation analysis applications. Remote Sensing of Environment, 91: 354-376.

Thompson, J.F., Stafford, J.V., and Miller P.C.H. (1991) Potential for automatic weed detection and selective herbicide application. Crop Protection, 10: 254-259.

Tian, L., Reid, J.F., and Hummel, J. (1999) Development of a precision sprayer for site- specific weed management. Transactions of the ASAE, 42(4): 893-900.

Vincent, L., and Soille, P. (1991) Watersheds in digital spaces: An efficient algorithm based on immersion simulations. IEEE Transactions on Pattern Analysis and

Machine Intelligence, 13(6): 583-598.

Vrindts, E., DeBaerdemaeker, J., and Ramon, H. (1999) Weed detection using canopy reflectance. In: Precision Agriculture 1999 2nd European Conference on Precision

Agriculture (JV Stafford, Ed.), Sheffield Academic Press, Sheffield, UK. pp. 257-

264.

Vrindts, E., DeBaerdemaeker, J., and Ramon, H. (2002) Weed detection using canopy reflection. Precision Agriculture. 3: 63-80.

Yang, C-C., Prasher, S., and Landry, J. (1998) Application of artificial neural networks to image recognition in precision agriculture. ASAE paper No. 98-3039.

Yang, C-C., Prasher, S., Landry, J-A., Ramaswamy, H.S., and Ditommaso, A. (2000) Application of artificial neural networks in image recognition and classification of crop and weeds. Canadian Agricultural Engineering 42(3): 147-152.

Yang, C-C., Prasher, S.O., Landry, J-A., and Ramaswamy, H.S. (2003) Development of a herbicide application map using artificial neural networks and fuzzy logic.

Agricultural Systems, 76: 561-574.

Zwiggelaar, R. (1998) A review of spectral properties of plants and their potential use for crop/weed discrimination in row crops. Crop Protection, 13(3): 189-206.

APPENDIX A

8° Hemispherical spectral reflectance coefficients for 12” x 12” Spectralon panel (SRT- 99-120).

Wavelength (nm) Reflectance Coefficient

400 .987

450 .987

500 .988

550 .988

600 .988

650 .987

700 .987

750 .987

800 .990

850 .989